Pay equity means paying people fairly for equal work: regardless of gender, race, age, or other protected characteristics. That sounds simple. In practice, it's one of the harder problems in HR because pay accumulates over years of individual decisions: different starting salaries, different raise amounts, different promotion timing. By the time a gap is visible, it's usually been building for a decade. See how Confirm handles performance management.
This guide covers pay equity analysis from scratch: how to define pay equity, how to conduct a pay equity audit, what to do when you find gaps, and how to build processes that prevent gaps from re-forming.
Who this is for: HR leaders, compensation analysts, and legal/compliance teams running their first (or fifth) pay equity audit. Covers both the methodology and the organizational change needed to make the results stick.
What Is Pay Equity?
Pay equity is the principle that employees should receive equal pay for equal work or work of equal value: without regard to gender, race, ethnicity, age, disability, or other protected characteristics.
There are two types of pay equity analysis, and they answer different questions:
| Type | What It Measures | Also Called |
|---|---|---|
| Unadjusted gap | Raw difference in median pay between groups (e.g., women earn 82 cents for every dollar men earn) | "Raw gap," "headline gap" |
| Adjusted gap | Difference after controlling for role, level, experience, and performance: the "apples to apples" comparison | "Controlled gap," "like-for-like gap" |
The adjusted gap is the legally relevant measure. A large unadjusted gap might be explained by representation differences (fewer women in senior roles), which is a pipeline problem, not a pay problem. An adjusted gap that can't be explained by legitimate factors is the actual pay equity issue.
Both matter. The unadjusted gap reflects the real outcome for employees. The adjusted gap tells you whether your pay decisions are fair.
Why Pay Equity Matters
The legal and business case are both strong:
Legal Risk
The Equal Pay Act (1963), Title VII, and state-level laws (California, Colorado, New York, and others) all require equal pay for equal work. Pay equity laws are getting stricter: Colorado, California, Washington, and New York now require posting salary ranges in job postings. The EU Pay Transparency Directive requires large employers to publish pay gap data. If you haven't audited your pay equity, you don't know your legal exposure.
Retention and Recruitment
When employees discover pay gaps: and they will, especially under transparency laws: the damage is severe. Employees who feel underpaid relative to peers don't just look for another job. They disengage first, then leave. The cost of replacing one employee is typically 50%–200% of annual salary. Preventing one departure from a pay equity finding more than pays for the audit.
Culture and Belonging
Employees who believe they're paid fairly are more engaged, more productive, and more likely to advocate for the company. Pay inequity: when discovered: destroys psychological safety in ways that no amount of culture programming can fix. It's a signal that the stated values and the actual practices don't match.
How to Conduct a Pay Equity Audit
Step 1: Define the Scope
Before pulling any data, decide what you're analyzing:
- Which employee population? All employees? US only? Full-time only? Include contractors?
- What pay elements? Base salary only, or total cash (base + bonus)? Equity grants?
- Which demographic groups? Gender is standard; race/ethnicity is increasingly included; disability, veteran status, and age are more complex
- What time period? Current snapshot or trend over 3–5 years?
Start with base salary for all full-time employees by gender and race/ethnicity. That's the standard audit scope. Expand from there once you've established the baseline.
Step 2: Build Your Dataset
Pull from your HRIS and compensation system. You'll need, for every employee:
- Employee ID (anonymize before analysis if sharing with external reviewers)
- Current base salary
- Job code and job family
- Pay grade/level
- Years at company
- Years in current role
- Most recent performance rating
- Location (city/state)
- Full-time or part-time status
- Gender
- Race/ethnicity (if collected)
Step 3: Run the Statistical Analysis
The standard methodology is multiple linear regression. You're modeling salary as a function of legitimate pay factors and then testing whether gender or race explains additional variation after those factors are controlled.
The model:
- Dependent variable: Log of base salary (log transformation normalizes the distribution)
- Control variables: Job level, job family, tenure, performance rating, location
- Test variables: Gender indicator, race/ethnicity indicators
If the coefficients on gender or race are statistically significant: meaning they explain pay differences that the control variables don't: you have an unexplained gap.
Practically, this means: after accounting for the fact that a Senior Engineer earns more than a Junior Engineer, and a 10-year employee earns more than a 2-year employee, are there still pay differences between men and women or between White and non-White employees at the same level with the same tenure and performance? If yes, that's the gap to address.
What "statistically significant" means here: A gap of less than $1,000 might not be statistically significant in a small dataset. A pattern of smaller gaps across hundreds of employees absolutely is. Both matter. Statistical significance tells you whether you can confidently attribute the gap to a systematic factor rather than random variation.
Step 4: Identify and Investigate Gaps
The regression gives you the aggregate picture. Now go to the individual level.
For each employee flagged as underpaid (typically anyone more than 10% below their predicted salary based on the model), review:
- When were they hired and at what salary? Was there a low starting offer that was never corrected?
- Have they received below-average merit increases? Why?
- Have they been passed over for promotions? Why?
- Is there a legitimate pay factor the model didn't capture (specialized certification, unique role scope)?
This qualitative review is where the real story comes out. The regression tells you something is wrong. The individual review tells you where it went wrong and why.
Step 5: Remediate
For employees with unexplained pay gaps:
- Calculate the remediation amount: what raise brings them to parity with similarly situated peers?
- Prioritize the largest gaps: especially employees who are both significantly underpaid and at retention risk
- Budget the correction: remediation is typically 0.1%–0.5% of payroll; model the cost before setting targets
- Implement in the current comp cycle: don't make employees wait for a future cycle when you've identified a problem
- Document the decisions: you'll need this for legal protection and for tracking whether the gaps close
There's no universal legal requirement to disclose the reason for a specific salary adjustment. Most organizations frame equity adjustments as standard annual review outcomes. What matters more: fixing the gap, documenting why the adjustment was made, and ensuring the manager communicates the change clearly. An unexplained raise is almost as confusing as no raise.
Step 6: Prevent Gaps from Re-forming
Remediation fixes today's gaps. It doesn't prevent new ones from forming. The structural changes that matter:
- Set salary ranges before starting a hire: not after you know the candidate's current salary
- Don't ask for salary history: in many states this is illegal; in all states it perpetuates historical inequities
- Apply starting salary rules consistently: anchoring all new hires at 90%–100% of midpoint, with documented exceptions for specialized experience
- Calibrate merit increases: review merit increase distributions by demographic group before finalizing to catch disparate impact before it becomes a gap
- Audit annually: not every five years. Once is not enough.
Pay Equity Audit Checklist
| Phase | Checklist Item | Status |
|---|---|---|
| Scope | Employee population defined | ☐ |
| Scope | Pay elements defined (base, total cash, equity) | ☐ |
| Scope | Demographic groups defined | ☐ |
| Data | Data extracted from HRIS | ☐ |
| Data | Missing values reviewed and resolved | ☐ |
| Data | Job codes verified for consistency | ☐ |
| Analysis | Unadjusted gap calculated | ☐ |
| Analysis | Regression model run with control variables | ☐ |
| Analysis | Individual employees flagged for review | ☐ |
| Investigation | Each flagged employee reviewed with manager | ☐ |
| Investigation | Legitimate factors documented or ruled out | ☐ |
| Remediation | Remediation amounts calculated | ☐ |
| Remediation | Budget approved | ☐ |
| Remediation | Adjustments implemented | ☐ |
| Prevention | Salary history ban in place | ☐ |
| Prevention | Starting salary guidelines documented | ☐ |
| Prevention | Annual audit cadence scheduled | ☐ |
Common Pay Equity Mistakes
"We fixed it once, we're good"
Pay equity is not a one-time project. Gaps re-form through new hire salary negotiations, merit increase disparities, and promotion timing differences. Annual audits are the minimum; quarterly monitoring for large organizations is better.
Only analyzing base salary
If your bonus payouts, equity grants, or commission structures have demographic disparities, the base salary analysis will look clean while the actual compensation gap is large. Audit total cash and total compensation, not merely base.
Using small comparison groups
If you have only 3 women in a particular job family and 15 men, statistical analysis won't give you a reliable result. Group jobs at a higher level of aggregation (e.g., all engineering roles at L4) when individual job families are too small.
Skipping the qualitative review
Regression finds patterns. It doesn't explain them. Two employees at the same level with the same tenure might have genuinely different pay for legitimate reasons (one was hired externally at a market premium; one completed a highly specialized certification). The individual review separates those cases from genuine equity issues.
Pay Equity and Pay Transparency
Pay transparency laws in California, Colorado, New York, Washington, and elsewhere require posting salary ranges in job listings. The EU Pay Transparency Directive (effective 2026 for large employers) goes further, requiring companies to publish aggregate pay gap data and provide employees access to pay range information.
Transparency accelerates equity. When employees can see pay ranges, they know when they're underpaid. HR teams that haven't done equity work before posting salary ranges are setting themselves up for an immediate credibility crisis.
The right sequence: audit and remediate first, then publish ranges. Not the other way around.
Frequently Asked Questions
How long does a pay equity audit take?
For a company of 200–500 employees with clean data, the analysis phase takes 2–4 weeks. Include 2–4 weeks for individual employee reviews, 2–4 weeks for budget approval and remediation planning. Total: 6–12 weeks from start to completed remediations. Larger companies with messy data can take 3–6 months.
Do we need an outside firm to do this?
Not necessarily. For companies under 500 employees with in-house HR analytics capability, a self-directed audit is feasible. Outside firms add value primarily for: legal privilege (if you involve employment counsel, the findings may be privileged), independence (harder to dismiss findings if they come from a third party), and statistical expertise. If legal exposure is your primary concern, involve counsel early.
What if we find a large gap we can't afford to fix immediately?
Document the finding and create a multi-year remediation plan. Courts and regulators look more favorably on companies that identified gaps and have a plan to close them than on companies that didn't look at all. Even if you can only fix 30% of identified gaps this year, documenting the commitment and the timeline is better than inaction.
How does pay equity analysis connect to performance management?
Closely. If performance ratings are biased: if managers systematically rate certain demographic groups lower: then merit increases based on those ratings will create or widen pay gaps. Performance calibration is a critical upstream control for pay equity. See Compensation Planning Guide for how to integrate the two processes.
Want to see how Confirm handles this? Request a demo — we'll walk you through the platform in 30 minutes.
If you're looking for calibration software to standardize ratings across your organization, see how Confirm approaches it.
